-
Kaleido-BERT: Vision-Language Pre-training on Fashion Domain, CVPR2021
-
**Describe the bug**
I tried to run the provided pascal VOC script using Apex optimization 01 and everything same as script except i was using a single GPU and hence changed the batch size to 24.
B…
-
Hi, Quanpeng Song. The author gives no clues about the net structure of style transformer, so I am wondering how you design the style transformer in your code and why it can work. Can you share someth…
-
Hello @arthurdouillard, I enjoyed reading your paper. I have a few queries regarding the implementation using the cityscapes dataset.
1. I am reproducing the experiments for [plop_cityscapesDomain_1-…
-
The results of DBSR reported in your paper are slightly higher than that reported in the CVPR2021's paper. Do you re-train the models or use different ways to calculate the metrics?
-
We would like to implement the Adversarial Laser Beam attack to simulate and investigate easily realisable physical attack.
Paper link: https://openaccess.thecvf.com/content/CVPR2021/papers/Duan_Ad…
-
Hi. In mean teacher, the consistency weight is 100. but in this work, all consistency weight is 1. Isn't this value too small?
Can you tell me the details of setting this parameter, because I have se…
-
-
Currently, in [Q4_0](https://github.com/ggerganov/ggml/pull/27) quantization we choose the scaling factor for each 32 group of weights as `abs(max(x_i))/7`. It is easy to see that this is suboptimal.
…
-
In the 4.2 Implementation, There is a sentence "We provide more details in the Appendix". But I can't find any Appendix in this paper. Would you please provide the appendix? I would appreciate it so m…